A non-Smad mechanism of fibroblast activation by transforming growth factor-beta via c-Abl and Egr-1: selective modulation by imatinib mesylate

Abstract

The nonreceptor protein tyrosine kinase c-Abl regulates cell proliferation and survival. Recent studies provide evidence that implicate c-Abl as a mediator for fibrotic responses induced by transforming growth factor-beta (TGF-beta), but the precise mechanisms underlying this novel oncogene function are unknown. Here, we report that when expressed in normal fibroblasts, a constitutively active mutant of Abl that causes chronic myelogenous leukemia (CML) stimulated the expression and transcriptional activity of the early growth response factor 1 (Egr-1). Mouse embryonic fibroblasts (MEFs), lacking c-Abl, were resistant to TGF-beta stimulation. Responsiveness of these MEFs to TGF-beta could be rescued by wild-type c-Abl, but not by a kinase-deficient mutant form of c-Abl. Furthermore, Abl kinase activity was necessary for the induction of Egr-1 by TGF-beta in normal fibroblasts, and Egr-1 was required for stimulation of collagen by Bcr-Abl. Lesional skin fibroblasts in mice with bleomycin-induced fibrosis of skin displayed evidence of c-Abl activation in situ, and elevated phospho-c-Abl correlated with increased local expression of Egr-1. Collectively, these results position Egr-1 downstream of c-Abl in the fibrotic response, delineate a novel Egr-1-dependent intracellular signaling mechanism that underlies the involvement of c-Abl in certain TGF-beta responses, and identify Egr-1 as a target of inhibition by imatinib. Furthermore, the findings show in situ activation of c-Abl paralleling the upregulated tissue expression of Egr-1 that accompanies fibrosis. Pharmacological targeting of c-Abl and its downstream effector pathways may, therefore, represent a novel therapeutic approach to blocking TGF-beta-dependent fibrotic processes.

Figure 1. Constitutively active Bcr-Abl stimulates collagen gene expression

Confluent foreskin fibroblasts were transiently transfected with Bcr-Abl (A, top panel; B) or wildtype c-Abl (A, lower panel) and incubated with TGF-β for 24 h. A. Whole cell lysates were subjected to Western analysis. Representative immunoblots are shown. B. Fibroblasts were cotransfected with 772COL1A2-CAT. Cell lysates were assayed for their CAT activities. Results, normalized with Renilla luciferase, are expressed as means ± S.D. of triplicate determinations. Open boxes, untreated fibroblasts; closed boxes, TGF-ß-treated fibroblasts. *p<0.005.

Figure 2. Imatinib blocks stimulation of collagen gene expression induced by Bcr-Abl or TGF-ß

Foreskin fibroblasts were pretreated with 10 μM imatinib (IM) for 30 min, followed by TGF-ß1 for a further 24 h. A. Whole cell lysates were subjected to Western analysis. Representative immunoblots. B. Total RNA was examined by real-time qPCR. Results normalized with actin are shown as means ± SD from triplicate determinations. Open boxes, untreated fibroblasts; closed boxes, TGF-ß-treated fibroblasts. *p<0.005. C. Fibroblasts were transiently transfected with 772COL1A2-CAT. Cell lysates were assayed for their CAT activities. The results are the means ± S.D. of triplicate determinations. Open boxes, untreated fibroblasts; closed boxes, TGF-ß-treated fibroblasts. D. Fibroblasts were transfected with Bcr-Abl and total RNA was subjected to realtime qPCR analysis. The results are representative of two independent experiments. * p<0.005. E. Whole cell lysates were subjected to Western analysis. Representative immunoblots are shown. IM, imatinib mesylate.

Figure 2. Imatinib blocks stimulation of collagen gene expression induced by Bcr-Abl or TGF-ß

Foreskin fibroblasts were pretreated with 10 μM imatinib (IM) for 30 min, followed by TGF-ß1 for a further 24 h. A. Whole cell lysates were subjected to Western analysis. Representative immunoblots. B. Total RNA was examined by real-time qPCR. Results normalized with actin are shown as means ± SD from triplicate determinations. Open boxes, untreated fibroblasts; closed boxes, TGF-ß-treated fibroblasts. *p<0.005. C. Fibroblasts were transiently transfected with 772COL1A2-CAT. Cell lysates were assayed for their CAT activities. The results are the means ± S.D. of triplicate determinations. Open boxes, untreated fibroblasts; closed boxes, TGF-ß-treated fibroblasts. D. Fibroblasts were transfected with Bcr-Abl and total RNA was subjected to realtime qPCR analysis. The results are representative of two independent experiments. * p<0.005. E. Whole cell lysates were subjected to Western analysis. Representative immunoblots are shown. IM, imatinib mesylate.

Figure 3. Impaired TGF-ß response in Abl^−/− Arg^−/− mouse embryonic fibroblasts

Fibroblasts from Abl ^−/−Arg ^−/− and wildtype mouse embryos in parallel were incubated with TGF-ß for 24 h. A. Whole cell lysates were subjected to Western analysis. Representative immunoblots. B. Following radiolabeling of cultures with [¹⁴C]-proline for 24 h, conditioned media were harvested and subjected to SDS-PAGE. Representative autoradiographs. C. Total RNA was subjected to real-time qPCR analysis. Results, expressed relative to actin, are shown as the means ± S.D.of triplicate determinations from a representative experiment. Open boxes, untreated fibroblasts; closed boxes, TGF-ß-treated fibroblasts. *p<0.005. D. Fibroblasts were transiently transfected with 772COL1A2-CAT. Following incubation with TGF-ß for 24 h, cell lysates were assayed for their CAT activities. The results are shown as the means ± S.D. of triplicate determinations. Open boxes, untreated fibroblasts; closed boxes, TGF-ß-treated fibroblasts. *p<0.005.

Figure 3. Impaired TGF-ß response in Abl^−/− Arg^−/− mouse embryonic fibroblasts

Fibroblasts from Abl ^−/−Arg ^−/− and wildtype mouse embryos in parallel were incubated with TGF-ß for 24 h. A. Whole cell lysates were subjected to Western analysis. Representative immunoblots. B. Following radiolabeling of cultures with [¹⁴C]-proline for 24 h, conditioned media were harvested and subjected to SDS-PAGE. Representative autoradiographs. C. Total RNA was subjected to real-time qPCR analysis. Results, expressed relative to actin, are shown as the means ± S.D.of triplicate determinations from a representative experiment. Open boxes, untreated fibroblasts; closed boxes, TGF-ß-treated fibroblasts. *p<0.005. D. Fibroblasts were transiently transfected with 772COL1A2-CAT. Following incubation with TGF-ß for 24 h, cell lysates were assayed for their CAT activities. The results are shown as the means ± S.D. of triplicate determinations. Open boxes, untreated fibroblasts; closed boxes, TGF-ß-treated fibroblasts. *p<0.005.

Figure 4. c-Abl rescues TGF-ß responses in Abl^−/− Arg^−/− MEFs

Abl ^−/−Arg ^−/− MEFs stably expressing c-Abl or a kinase-inactive mutant form of cAbl (KD) in parallel were incubated with TGF-ß for 24 h. A. Whole cell lysates were subjected to Western analysis. Representative immunoblots. B. Total RNA was subjected to real-time qPCR analysis. Results, expressed relative to actin are means ± S.D. from triplicate determinations from a representative experiment. Open boxes, untreated fibroblasts; closed boxes, TGF-ß-treated fibroblasts. *p<0.005.

Figure 5. TGF-ß induces c-Abl kinase activity

A. Confluent fibroblasts were incubated with TGF-ß1 (10 ng/ml) or PDGF (25 ng/ml) for indicated periods. Whole cell lysates were immunoprecipitated with antibodies to c-Abl and kinase activity was assayed using GST-Crk as substrate (upper panel). Western analysis of whole cell lysates (lower panel). B. Fibroblasts were incubated with TGF-β1 for 30 min and whole cell lysates were subjected to Western analysis.

Figure 6. Smad2/3-independent inhibition of TGF-ß responses by imatinib

A-C. Confluent foreskin fibroblasts were pretreated for 30 min with imatinib (10 μM) followed by incubation with TGF-ß1 for a further 30 min. A. Whole cell lysates were subjected to Western analysis. Representative immunoblots. B. Nuclear extracts were subjected to DNA affinity precipitation assays using biotin end-labeled oligonucleotides harboring the SBE sequence as described under “Materials and Methods.” A representative Western blot. C . Fibroblasts were immunostained with antibodies to Smad4, and examined by immunofluorescence confocal microscopy. Representative images. IM, Imatinib mesylate. D. Abl ^−/−/Arg ^−/− MEFs and wildtype MEFs in parallel were incubated with TGF-ß1 (30 min) and Smad2/3 localization was examined by confocal microscopy. Representative images (left panel). The nuclear localization of Smad2/3 was quantitated as described under methods (right panel). Results represents the means ± S.D. from separate fields. Open boxes, untreated fibroblasts; closed boxes, TGF-ß-treated fibroblasts. *p<0.005.

Figure 7. Bcr-Abl stimulates Egr-1

A. Confluent foreskin fibroblasts were transiently transfected with Bcr-Abl and incubated with TGF-β for 24 h. A. Whole cell lysates were subjected to Western analysis. Representative immunoblots are shown. B. Confluent fibroblasts were transiently transfected with Bcr-Abl in presence of imatinib. Total RNA normalized with actin was subjected to realtime qPCR analysis. The results are the means ± S.D. of triplicate determinations. *p<0.005. C. Whole cell lysates from confluent cultures treated with imatinib were subjected to Western analysis. Representative immunoblots are shown.

Figure 8. Egr-1 acts downstream of c-Abl in TGF-ß signaling

A. NIH3T3 fibroblasts were transiently cotransfected with indicated concentrations of Bcr-Abl along with Egr-1-luc. Following pretreatment with imatinib or U0126, cultures were incubated with TGF-ß1 for a further 24 h and cell lysates were assayed for their luciferase activities. The results, normalized with Renilla luciferase, are expressed as fold-induction relative to TGF-ß or Bcr-Abl. *p<0.005. B. Abl ^−/−Arg ^−/− MEFs stably expressing c-Abl or a kinase-deficient mutant form of c-Abl (KD) were transfected with Egr-1- luc, and following incubation with TGF-ß1 for 24 h, cell lysates were assayed for their luciferase activities. The results, normalized with Renilla luciferase, are expressed as means ± S.D. of triplicate determinations from two independent experiments. Open boxes, untreated MEFs; closed boxes, TGF-ß-treated MEFs. C. Wildtype and Egr-1^−/− skin fibroblasts were transfected with Bcr-Abl and total RNA were subjected to realtime qPCR analysis (upper panel), and secreted Type I collagen was examined by Western blot analysis (lower panel). The results are the means ± S.D. of triplicate determinations. D. ChIP assays. Foreskin fibroblasts transfected with Bcr-Abl were cross-linked with formaledehyde, chromatin was immunoprecipitated with antibodies to Egr-1, followed by PCR amplification with COL1A2-specific primers. PCR products were analyzed by agarose gel electrophoresis. Input genomic DNA was used as positive control. IM, Imatinib mesylate.

Figure 8. Egr-1 acts downstream of c-Abl in TGF-ß signaling

A. NIH3T3 fibroblasts were transiently cotransfected with indicated concentrations of Bcr-Abl along with Egr-1-luc. Following pretreatment with imatinib or U0126, cultures were incubated with TGF-ß1 for a further 24 h and cell lysates were assayed for their luciferase activities. The results, normalized with Renilla luciferase, are expressed as fold-induction relative to TGF-ß or Bcr-Abl. *p<0.005. B. Abl ^−/−Arg ^−/− MEFs stably expressing c-Abl or a kinase-deficient mutant form of c-Abl (KD) were transfected with Egr-1- luc, and following incubation with TGF-ß1 for 24 h, cell lysates were assayed for their luciferase activities. The results, normalized with Renilla luciferase, are expressed as means ± S.D. of triplicate determinations from two independent experiments. Open boxes, untreated MEFs; closed boxes, TGF-ß-treated MEFs. C. Wildtype and Egr-1^−/− skin fibroblasts were transfected with Bcr-Abl and total RNA were subjected to realtime qPCR analysis (upper panel), and secreted Type I collagen was examined by Western blot analysis (lower panel). The results are the means ± S.D. of triplicate determinations. D. ChIP assays. Foreskin fibroblasts transfected with Bcr-Abl were cross-linked with formaledehyde, chromatin was immunoprecipitated with antibodies to Egr-1, followed by PCR amplification with COL1A2-specific primers. PCR products were analyzed by agarose gel electrophoresis. Input genomic DNA was used as positive control. IM, Imatinib mesylate.

Figure 9. Increased expression of phospho-cAbl and Egr-1 in lesional tissue in bleomycin-induced scleroderma in the mouse

6-weeks old BALB/C mice received daily s.c. injections of bleomycin (BLM) or PBS. At day 28 after injections were started, lesional skin was examined by immunohistochemistry using antibodies against phospho-cAbl or Egr-1. Representative photomicrographs are shown. Original magnification × 400 (upper panels) or × 1000 (lower panels). The proportion of fibroblasts in the lesional dermis immunopositive for phospho-cAbl was determined. Results represents the means ± S.D. from six separate fields from at least three mice/group. Human breast carcinoma tissue was used as positive control.